Static Polarizabilities of Dielectric Nanoclusters

TL;DR

This paper presents a microscopic method to accurately evaluate the static polarizability of dielectric nanoclusters, highlighting the limitations of simple additive models and the importance of surface effects.

Contribution

It introduces an exact microscopic approach for calculating cluster polarizabilities and compares it with additive and continuum models, emphasizing surface effects.

Findings

Additivity approximation often significantly underestimates or overestimates polarizability.

Surface effects are substantial near edges and surfaces.

Continuum models can be less accurate for small or irregular clusters.

Abstract

A cluster consisting of many atoms or molecules may be considered, in some circustances, to be a single large molecule with a well defined polarizability. Once the polarizability of such a cluster is known, one can evaluate certain properties, e.g. the cluster's van der Waals interactions, using expressions derived for molecules. In the present work, we evaluate the static polarizability of a cluster using a microscopic method that is exact within the linear and dipolar approximations. Numerical examples are presented for various shapes and sizes of clusters composed of identical atoms, where the term "atom" actually refers to a generic constituent, which could be any polarizable entity. The results for the cluster's polarizabilities are compared with those obtained by assuming simple additivity of the constituents' atomic polarizabilities; in many cases, the difference is large, demonstrating the inadequacy of the additivity approximation. Comparison is made (for symmetrical geometries) with results obtained from continuum models of the polarizability. Also, the surface effects due to the nonuniform local field near a surface or edge are shown to be significant.